Reinforced Learning in Children through a Stress Warning Unit

Stress is one of the main factors that interrupts learning in children. Such stress can be activated by environment or daily activities especially at the primary school. Children must adapt themselves to learn new subjects and contents along with difficult assignments that could lead to an occurrence of stress in children. When stress occurs, the body releases cortisol which negatively affect the cortex (involves in learning and thinking processes) and hippocampus (involved in the development of intelligence and emotion). This research thus aims to develop a wearable device to measure stress levels in children. This would assist parents and teachers to notice a change in stress during children' study or performing any activities. To estimate stress levels, we processed the pulse rate variability (PRV) which extracted from photoplethysmogram (PPG) signals. We employed the optical reflection sensors to obtain the PPG signals on the wrist area. The ESP32 microcontroller was programmed to acquire and process such PPG signals. Quantitative stress levels were then calculated from the average distinct frequency (f
_d
) of the PRV in the range of 0.04 - 0.15 Hz which reflects by the parasympathetic and sympathetic nervous systems when perceived stress. In our experiments, we conducted the standard trier social stress test whereby cognitive stress periods (an arithmetic test) and a relax period were performed while wearing our device. We then calculated f
_d
in a particular period of 10 healthy participants. It was found that stress levels can be divided into 4 groups based on a calculation of f
_d
ratio (%): (i) no stressor 0 - 25%, (ii) low stress or 25 - 50%, (iii) moderate stress or 50 - 75% and (iv) high stress or 75 - 100%. These results indicate a potential to monitor stress in children during learning activities efficiently and would possibly enable a self-awareness for coping stress in children.